AMICI: Attention Mechanism Interpretation of Cell-cell Interactions

Cross-attention-based cell-cell interaction inference from ST data.

Installation

You need to have Python 3.10 or newer installed on your system. If you don't have Python installed, we recommend installing Mambaforge.

1. Install the latest release of amici-st from PyPI <https://pypi.org/project/amici-st/>_:

pip install amici-st

Or install the latest development version via the following command:

pip install git+https://github.com/azizilab/amici.git@main

2. Import amici

import anndata
from amici import AMICI

adata = anndata.read_h5ad("./adata.h5ad")
AMICI.setup_anndata(adata, labels_key="cell_type", coord_obsm_key="spatial")
model = AMICI(adata, **model_params)
model.train()

Documentation

Find more detailed documentation on AMICI here: AMICI documentation.

To get started, check out our tutorial on basic usage of AMICI here: Basic Usage Tutorial.

Citation

If you find our work useful, please cite our preprint: https://www.biorxiv.org/content/10.1101/2025.09.22.677860v1

AMICI: Attention Mechanism Interpretation of Cell-cell Interactions

@article{Hong2025.09.22.677860,
    title = {AMICI: Attention Mechanism Interpretation of Cell-cell Interactions},
    author = {Hong, Justin and Desai, Khushi and Nguyen, Tu Duyen and Nazaret, Achille and Levy, Nathan and Ergen, Can and Plitas, George and Azizi, Elham},
    doi = {10.1101/2025.09.22.677860},
	journal = {bioRxiv},
	publisher = {Cold Spring Harbor Laboratory},
	year = {2025},
}

Disclaimer

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Justin Hong, Khushi Desai, and Elham Azizi are inventors on a provisional patent application having U.S. Serial No. 63/884,704, filed on September 19, 2025, by The Trustees of Columbia University in the City of New York directed to the subject matter of the manuscript associated with this repository.